A Bioinformatics Approach to Establishing Novel Transforming Growth Factor-Beta Target Genes in Glioblastoma.

Abstract

Abstract Introduction TGFβ inhibition is becoming an increasingly exciting prospect for the treatment of glioblastoma (GBM) as more about its efficacy is understood. However, there is little demonstrated improvement in patient outcomes from clinical trials using small molecule inhibitors of the TGFβ receptors. Clearly, the potential of TGFβ therapy has not yet been realised. We scrutinised the TGFβ canonical genes at a cytogenetic, genetic and gene expression level for novel therapeutic targets. Materials and Methods 10 canonical TGFβ genes were analysed. Karyotypes were interpreted using the International System for Human Cytogenetic Nomenclature and graphics generated using the Cytogenetic Data Analysis System. Analysis and generation of graphics were carried out using The Cancer Genome Atlas (TCGA) Provisional Glioblastoma set of 290 whole-exome-sequenced tumours via cBioPortal. mRNA expression data were extracted from The Cancer Genome Atlas (TCGA) via cBioPortal. Kaplan-Meier graphs were generated using GraphPad Prism®, for which pairwise log rank comparisons were conducted to determine a difference in both overall survival (OS) and disease-free, progression-free (DFPF) survival. Results At cytogenetic and genetic levels, TGFβ genes are not markedly altered in GBM. However, mRNA expression of TGFβ2 is enhanced in GBM compared to other cancers. In addition, we correlated upregulation of both SMAD7 and SARA with improved OS and upregulation of SARA with improved DFPF survival. Conclusions TGFβ2 upregulation may be associated with GBM tumourigenesis and SMAD7 and SARA upregulation may be associated with improved survival. We demonstrate that three canonical genes have therapeutic potential, but their conflicting roles must be further understood before TGFβ-modulating therapies are successful.